Developers often examine a failed program's state to diagnose and fix software bugs. For this reason, operating systems and programming suites include tools to capture a program's state in a core file at crash time. The recent advent of ubiquitous network connectivity for personal computers makes remote crash reporting possible, whereby programs send crash information back to developers after a failure. This practice, which allows developers to receive information about bugs in their programs after the programs have been distributed to users, has since become commonplace [1,2,3,4]. Remote crash reporting offers many readily apparent benefits to developers, but the privacy-related implications of the technology are not as well understood.
The contributions of this paper include